Small world properties changes in mild traumatic brain injury

PurposeTo investigate local and global efficiency changes characterized by small-world properties based on resting-state functional MRI, such as centrality and clustering coefficient, in mild traumatic brain injury (MTBI) patients; and to associate these findings with axonal injury as measured by diffusion tensor imaging (DTI) as well as with post-concussive symptom (PCS). Materials and MethodsThirty patients (mean age 3513 years) with clinically defined MTBI and 45 age-matched healthy controls (mean age 3710 years) participated in the experiments. Resting-state functional MRI was performed using gradient echo planar imaging sequence with 3 Tesla MRI scanner to obtain functional small-world networks. Out of all participants, 20 MTBI patients and 20 controls had available DTI data with three b-values (0, 500, 1000) s/mm(2) and 30 directions for diffuse axonal injury analyses. ResultsCompared with controls, MTBI patients showed lower relative betweenness centrality (P=0.01), but significantly higher clustering coefficient (P=0.04), and these two metrics correlated negatively in patients (r=-0.77; P<0.001). Regions with lower betweenness centrality (e.g., frontal and occipital) corresponded with the regions of reduced FA in patients, while global FA reduction correlated with betweenness centrality (r=0.48; P=0.03) and clustering coefficient (r=-0.46; P=0.04) in MTBI patients. In addition, there was significantly higher thalamocortical connectivity that correlated with clustering coefficient (r=0.39; P=0.03) in patients. Also, patients with higher clustering coefficient tended to have less PCS score with negative correlation (r=-0.4; P=0.04). ConclusionOur results demonstrated significant functional small-world properties changes in patients with MTBI, and suggest decreased global efficiency, possibly due to diffuse axonal injury and local network upregulation including increased thalamo-cortical connectivity. Level of Evidence: 2 Technical Efficacy: Stage 2 J. MAGN. RESON. IMAGING 2017;46:518-527